Groundwater depletion in semi‑arid regions requires accurate forecasting and proactive management. This study integrates wavelet analysis, diagnostic SARIMA modeling, sensitivity analysis, comparative physical/machine learning models, and managed aquifer recharge (MAR) scenario evaluation to forecast groundwater levels in the Galedar aquifer, Iran (2006–2024). Wavelet decomposition reveals decadal variability (48.5% variance) in regional flow zones and annual dominance (45%) in the high‑recovery zone, with coherence to precipitation (1–3 month lag) and ENSO (r=0.35–0.50). Manual SARIMA achieves out‑of‑sample RMSE of 0.062–0.162 m. Sensitivity identifies specific yield as dominant (Sobol’ ST=0.58–0.71). The hybrid SARIMA‑wavelet model reduces RMSE by 16–31%. Under business‑as‑usual, the mean water level declines 2.6 m (0.26 m/year) by 2034, with cumulative storage loss (191.1 MCM) exceeding recoverable storage (150 MCM). MAR scenarios show: high‑recovery zone recharge (+1.41 m, 71% decline reduction); distributed recharge (1.8‑fold greater storage recovery); episodic recharge (+1.11 m, 40% lower efficiency). Cost‑benefit analysis confirms positive net present value (BCR 1.4–3.4). The framework supports risk‑based, spatially stratified management for stressed aquifers.
Somayeh Zarei Doudeji (Mon,) studied this question.